基于广义维数距离的语音端点检测方法

武薇; 范影乐; 庞全

doi:10.3724/SP.J.1146.2005.00963

基于广义维数距离的语音端点检测方法

doi: 10.3724/SP.J.1146.2005.00963

基金项目:

国家自然科学基金(60302027)和浙江省教育厅科研基金(20030620)资助课题

计量
- 文章访问数: 3302
- HTML全文浏览量: 95
- PDF下载量: 738
- 被引次数: 0
出版历程
- 收稿日期: 2005-08-01
- 修回日期: 2006-03-20
- 刊出日期: 2007-02-19

A Speech Endpoint Detection Method Based on the Feature Distance of Generalized Dimension

摘要

摘要: 为能够准确有效地对含噪声语音信号进行起止位置的端点检测，该文提出了一种基于广义维数距离的端点检测方法。首先利用覆盖法求取广义维数得到该语音信号的三维特征向量，包括容量维数、信息维数、关联维数；然后计算信号的维数特征距离；最后根据特征距离对语音信号类别进行决策分类。实验结果表明，与仅使用单一维数特征检测语音起止端点相比，该文所提出的方法具有较好的鲁棒性，对混杂有不同噪声、不同信噪比的语音信号都能有较好的检测结果，尤其适用于低信噪比的语音端点检测。
- 语音端点检测;广义维数;特征距离
Abstract: Based on the feature distance of generalized dimension, a speech endpoint detection method is proposed in order to detect the noisy-corrupted speech efficiently. Through calculating the generalized dimension by covering the signal with n-dimension boxes, three dimension feature vectors including the box dimension, the information dimension and the correlation dimension are got. Then dimension feature distance could be calculated and used to make a classification for the speech signal. Experimental results show that compared with the detection using one dimension feature only, the proposed method is more robust to the endpoint detection of speech signal containing different noise and SNR, especially for the lower SNR signal.

HTML全文

参考文献(1)

[1] 沈亚强. 低信噪比语音信号端点检测和自适应滤波. 电子测量与仪器学报, 2001, 15(1): 27-32. Shen Ya-qiang. Low SNR speech signal endpoints detection and adaptive filtering. Journal of Electronic Measurement and Instrument, 2001, 15(1): 27-32. [2] 陈国, 胡修林, 张蕴玉等. 汉语普通话语音的分形特性及其盒维数的统计分析. 信号处理, 2000, 16(12): 297-301. [3] Turiel A and Prez-Vicente C. Role of multifractal sources in the analysis of stock market time series[J].Physica A: Statistical Mechanics and its Applications.2005, 355(24):475-496 [4] 徐玉秀, 侯荣涛, 杨文平. 广义分形维数在旋转机械故障诊断中的应用研究. 中国机械工程, 2003, 21: 1812-1814. [5] F Ferens K and Kinsner W. Multifractal texture classification of images[J].WESCANEX 95. Communications, Power, and Computing. Conference Proceedings, IEEE. Winnipeg, Manitoba, Canada. May 15-1.1995, Vol.2:438-444 [6] Chaudhuri B B and Sarkar N. An efficient approach to compute fractal dimension in texture image. Pattern Recognition, Conference A: 11th IAPR International Conference on Computer Vision and Applications, Proceedings, The Hague, Netherlands, 30 Aug.-3 Sept., 1992, vol.1: 358-361. [7] Nugraha H B and Langi A Z R. Segmented fractal dimension measurement of 1-D signals: A wavelet based method[J].2002. APCCAS '02. 2002 Asia-Pacific Conference on Circuits and Systems, Bali, Indonesia.2002, vol.1:195-198 [8] Grieder W and Kinsner W. Speech segmentation by variance fractal dimension. 1994, Conference Proceedings. 1994 Canadian Conference on Electrical and Computer Engineering, Halifax, Canada, 25-28 Sept. 1994, vol.2: 481-485. [9] Boshoff H F V. A fast box counting algorithm for determining the fractal dimension of sampled continuous functions. 1992. COMSIG '92, Proceedings of the 1992 South African Symposium on Communications and Signal Processing, New York, NY, USA, 11 Sept. 1992: 43-48. [10] 边肇祺, 张学工编著. 模式识别. 第2版, 北京: 清华大学出版社, 2000: 185-186. [11] Jia Chuan and Xu Bo. An improved entropy-based endpoint detection algorithm. International Symposium on Chinese Spoken Language Processing (ISCSLP 2002). Taipei, Taiwan. August 23-24, 2002: 479-583. [12] Lingyun Gu and Zahorian S. A new robust algorithm for isolated word endpoint detection. 2002. Proceedings. (ICASSP '02). IEEE International Conference on Acoustics, Speech, and Signal Processing, Orlando, Florida, USA, 13-17 May 2002, vol.4, IV-4161.

施引文献

资源附件(0)

访问统计